Bertrand Competition, explained.
Bertrand competition models firms setting prices, with demand allocation depending on those prices. For identical products in the simplest model, a lower-priced firm can capture the market.
Why it happens
At equal prices the firms split demand. A one-token undercut captures all demand in this teaching game, but cuts the margin per unit. Going below the cost of 20 can generate a loss even while capturing sales. The sharp switch in demand relies on consumers treating the products as interchangeable.
Bertrand competition models firms setting prices. For identical goods with equal constant costs and sufficient capacity, a lower price can capture demand from a higher-priced rival.
Read the result
Compare matching the fixed rival price with undercutting by one token and with a much larger cut. Submit each price and track profit rather than only sales. The rival does not retaliate here, so these rounds inspect one-shot responses rather than reproduce a price war or equilibrium convergence.
A worked example
Matching or undercutting a price of fifty
Both shops post 50. Demand is 50 units, split equally, so your profit is 25×30=750 teaching tokens.
Post 49 against the same fixed rival: demand becomes 51, all served by you, and profit is 51×29=1,479.
A much lower price can reduce profit even with more customers. At 10, selling 90 units loses 900 because the cost remains 20.
OPTIONAL DEEPER DETAILGo deeper: inside the model
Inside this model
Demand is max(0, 100 − the lowest price). The low-price seller serves it all; a tie splits it equally. Unit cost is 20 and profit is (own price − 20) × own units. Integer prices are allowed from 10 to 90; below-cost prices can generate losses. Rival price is held fixed for comparison.
Where this idea is useful
A practical use
An undifferentiated market can make a small price cut shift sales sharply, even though it also reduces the margin per unit.
A common misconception
“The seller with the most customers must be doing best.”
Sales volume ignores the margin and total costs. In this model a below-cost seller wins demand while losing tokens on every unit.
What this explanation leaves out
- This is a one-shot best-response challenge, not a simulation of convergence to Bertrand equilibrium. Branding, search costs, capacities, and repeated interaction can change the outcome.
Why do real firms often charge different prices without losing all sales?
Products can differ in quality, convenience, trust, location, and capacity. Search costs also matter. Those features are intentionally absent from the identical-product model.
Which feature makes your customers see competing products as different rather than interchangeable?
Associated thinkers
Associations marked provisional are awaiting source review.
Further reading
Explore the original research or the teaching reference behind this experiment.